Perplexity is an AI-powered search engine that answers questions in plain language and shows numbered citations to the web pages it used. Instead of scanning a list of links, users get a synthesized answer and can click through to verify any claim.
Because Perplexity cites sources so visibly, it is one of the easiest AI platforms for businesses to earn traffic and credibility from. This guide explains how it works, how its crawlers behave, and what you can do to become a source it trusts.
What is Perplexity?
Perplexity is an AI search engine and answer engine. According to its help center, it searches the web in real time when you ask a question, combines information from multiple sources into a direct answer, and includes citations and links back to the originals.
Perplexity offers several search modes, including regular Search, Pro Search and Deep Research, and lets paid users choose between models from providers such as OpenAI and Anthropic. Its product changes fast. Its changelog shows frequent releases through 2026, including agent-style "Computer" features, automations and new model options. Treat feature details as accurate as of October 2026.
Why Perplexity matters for marketers
Perplexity's design puts sources front and center. Citations appear as numbered references in the answer and as source cards, so being cited gives your brand visible credit and a direct path for clicks. That makes it a meaningful source of AI referral traffic, even if its audience is smaller than Google's.
Perplexity users also tend to ask research-style questions: comparisons, explanations, product evaluations. Those are exactly the moments when a well-cited page can shape a buying decision. And since Perplexity searches the live web for each question, fresh, well-structured content can be picked up without waiting for a model to be retrained.
How Perplexity works
Real-time retrieval
When you ask a question, Perplexity searches the web, gathers relevant pages, and passes the most useful passages to a large language model. This approach is a form of retrieval-augmented generation (RAG): the model writes its answer from retrieved documents rather than from memory alone.
Synthesis and citations
The model writes a conversational answer and attaches numbered citations to the claims, linking to the source pages. Users can ask follow-up questions, and Perplexity keeps the context of the thread.
Perplexity's crawlers
Perplexity documents two user agents, and the difference matters:
- PerplexityBot is designed to surface and link websites in Perplexity's search results. Perplexity says it is not used to crawl content for AI foundation models. To appear in Perplexity's results, you need to allow it in robots.txt.
- Perplexity-User fetches pages when a user's question requires it. Perplexity says it generally ignores robots.txt because the fetch is user-requested, and it is not used for training.
Perplexity also publishes IP address lists for both bots, so you can allow them in a web application firewall using a user-agent match plus an IP check.
How to get cited by Perplexity
- Allow PerplexityBot. Check robots.txt and your CDN or firewall. Many sites block AI bots by default through security settings without realizing it.
- Lead with the answer. Start pages and sections with a concise, self-contained answer to the question in the heading. Short, clear passages are easier to extract and cite.
- Be specific and factual. Concrete details like prices, steps, specs, dates and definitions give the model something precise to reference. Vague marketing copy rarely gets cited.
- Keep content current. Since Perplexity searches live, update key pages regularly and show clear update dates.
- Structure pages clearly. Use descriptive headings, short paragraphs and lists. Breaking content into focused sections, sometimes called content chunking, helps retrieval systems match the right passage to a question.
- Earn mentions on trusted sites. Perplexity's help center notes that answers draw on sources like news organizations, academic publications and established content sites. Being referenced on those sites gives you more paths into answers.
- Track it with prompt testing. Run a fixed set of real customer questions monthly and note which sources Perplexity cites.
Perplexity in practice
Picture a B2B software company selling inventory tools to small retailers. Someone asks Perplexity, "What's the easiest inventory software for a shop with two locations?" Perplexity may pull from review sites, comparison articles and vendor pages. If the company has a page that directly addresses multi-location inventory, with plain-language pricing and setup steps, and has been mentioned in a couple of independent roundups, it has more ways to be included. If its robots.txt or firewall blocks PerplexityBot, it has almost none.
Common mistakes and misconceptions
- Blocking every AI bot at once. A blanket block can remove you from Perplexity's results entirely. Decide per crawler based on what each one does.
- Assuming Perplexity uses your content for training. Perplexity says PerplexityBot isn't used to crawl for foundation models. Its purpose is search results.
- Relying on robots.txt to stop Perplexity-User. Perplexity says this agent generally ignores robots.txt because a user requested the page.
- Writing only for Google keywords. Perplexity questions are longer and more conversational. Content that answers the full question tends to fit better.
- Ignoring off-site presence. Your own pages are only part of the picture. Independent mentions help the model see you as a credible option.
How Butter approaches Perplexity
At Butter, Perplexity is one of the AI platforms we check every month in our GEO plan's prompt testing. Because it shows its citations so clearly, it's a useful way to see which pages and publications AI search trusts in your category. We combine that testing with indexing strategy, schema markup guidance, knowledge graph submissions, and one AI-crawlable content citation and backlink each month, then report what we delivered and what to fix on and off your site. More detail is on our GEO service page. We don't promise citations, but we work on the things that make them more likely.
Frequently asked questions
Is Perplexity a search engine or a chatbot?
It is both. Perplexity describes itself as an AI-powered search engine that searches the web in real time and returns a conversational answer with citations. You can ask follow-up questions like a chatbot.
How do I get my website cited by Perplexity?
Start by allowing PerplexityBot in robots.txt and your firewall. Then publish clear, specific, up-to-date content that directly answers questions, and earn mentions on trusted sites. No one can guarantee a citation.
Does Perplexity train its AI models on my content?
Perplexity says PerplexityBot is not used to crawl content for AI foundation models, and Perplexity-User is not used for training either. Its crawler is meant to surface and link sites in search results.